Upload
others
View
3
Download
0
Embed Size (px)
Citation preview
Evidence for “Underlying” XV Word Order in Early Old French
DiGS 16Budapest
Anthony Kroch and Beatrice SantoriniUniversity of Pennsylvania
Tuesday, July 8, 14
www.ling.upenn.edu/~kroch/handouts/digs16.pdf
Tuesday, July 8, 14
I. Statistical independence in word order patterns in Yiddish
(Santorini 1993)
Tuesday, July 8, 14
(1) ven der vatr nurt doyts leyen kan if the father only German read can
(2) ven der vatr leyen kan nurt doyts if the father read can only German
Noun Phrase Extraposition in Yiddish
Tuesday, July 8, 14
(3) dz ikh reyn fun der ashin verde that I clean from the ashes become
(4) dz ikh reyn verde fun der ashin that I clean become from the ashes
Prepositional Phrase Extraposition in Yiddish
Tuesday, July 8, 14
Date DP postposing PP postposingPostposed Not postposed freq. Postposed Not postposed freq.
1400-1489 1 12 .08 9 12 .431490-1539 7 19 .27 13 16 .451540-1589 7 24 .23 52 21 .711590-1639 10 40 .20 39 23 .631640-1689 4 19 .17 17 30 .361690-1739 1 5 .17 6 3 .671740-1789 1 2 .33 8 7 .531790-1839 0 1 .00 1 1 .501840-1950 no INFL-final data – no INFL-final data –
Table 5: Frequencies of DP and PP postposition by time period.
LSA.237: lecture 6 8
Frequency of DP and PP Postposingin the History of Yiddish (Santorini 1993)
Tuesday, July 8, 14
0
0.2
0.4
0.6
0.8
1445 1515 1565 1615 1665 1715 1765 1815 1840+
PP
DP
Graph of DP and PP Postposing in the History of Yiddish
Tuesday, July 8, 14
II. Statistical independence in word order patterns in Ancient Greek
(Taylor 1994)
Tuesday, July 8, 14
SVX: Καµβύσης ἐδέξατο τὰ δῶρα
VSX: ἐδέξατο Καµβύσης τὰ δῶρα
VXS: ἐδέξατο τὰ δῶρα Καµβύσης
XSV: τὰ δῶρα Καµβύσης ἐδέξατο
XVS: τὰ δῶρα ἐδέξατο Καµβύσης
SXV: Καµβύσης τὰ δῶρα ἐδέξατο Kambuses the gifts received
“Free” Word Order in Ancient Greek
Tuesday, July 8, 14
pattern formulaS , X v (1-s)(1-p)X , Y v (1-p)2
S v X p(1-s)X v S s(1-p)X v Y 2p(1-p)v S , X spv X , Y p2
Table 10: Formula for calculating distribution of clause types based on theprobability of postposing of subjects and complements
Language Change: lecture 4 21
pattern formulaS , X v (1-s)(1-p)X , Y v (1-p)2
S v X p(1-s)X v S s(1-p)X v Y 2p(1-p)v S , X spv X , Y p2
Table 10: Formula for calculating distribution of clause types based on theprobability of postposing of subjects and complements
Language Change: lecture 4 21
pattern formulaS , X v (1-s)(1-p)X , Y v (1-p)2
S v X p(1-s)X v S s(1-p)X v Y 2p(1-p)v S , X spv X , Y p2
Table 10: Formula for calculating distribution of clause types based on theprobability of postposing of subjects and complements
Language Change: lecture 4 21
Tuesday, July 8, 14
pattern formula observed distribution expected distributionS X v (1-s)(1-p) 41 41X Y v (1-p)2 7 5S v X p(1-s) 31 31X v S s(1-p) 13 12X v Y 2p(1-p) 4 8v S X sp 8 9v X Y p2 5 3Total N = 109 s = .23 p = .43N for clauses with subjects = 93N for clauses without subjects = 16�2 = 4.12, p > .8
Table 11: Best fit of observed and expected distribution of clauses with 2arguments in Homer
Language Change: lecture 4 22
pattern formula observed distribution expected distributionS X v (1-s)(1-p) 41 41X Y v (1-p)2 7 5S v X p(1-s) 31 31X v S s(1-p) 13 12X v Y 2p(1-p) 4 8v S X sp 8 9v X Y p2 5 3Total N = 109 s = .23 p = .43N for clauses with subjects = 93N for clauses without subjects = 16�2 = 4.12, p > .8
Table 11: Best fit of observed and expected distribution of clauses with 2arguments in Homer
Language Change: lecture 4 22
Tuesday, July 8, 14
pattern formula observed distribution expected distributionS X Y v (1-s)(1-p)2 4 5S X v Y 2p(1-s)(1-p) 10 7X Y v S s(1-p)2 2 2S v X Y p2(1-s) 1 3X v S Y 2p(1-p)(1-s) 4 3v S X Y s(p2) 0 1Total N = 21 s = .29 p = .41�2 = 3.66, p > .8
Table 12: Best fit of observed and expected distribution of clauses with 3arguments in Homer
Language Change: lecture 4 23
pattern formula observed distribution expected distributionS X Y v (1-s)(1-p)2 4 5S X v Y 2p(1-s)(1-p) 10 7X Y v S s(1-p)2 2 2S v X Y p2(1-s) 1 3X v S Y 2p(1-p)(1-s) 4 3v S X Y s(p2) 0 1Total N = 21 s = .29 p = .41�2 = 3.66, p > .8
Table 12: Best fit of observed and expected distribution of clauses with 3arguments in Homer
Language Change: lecture 4 23
Tuesday, July 8, 14
subject postposing NP compl. postposing N1 NP argument .27 .44 1122 NP arguments .23 .48 1093 NP arguments .29 .41 21
1 NP/1 PP argument .28 .48 58
Table 13: Probability of subject and NP complement postposing in fourHomeric data sets
Language Change: lecture 4 24
subject postposing NP compl. postposing N1 NP argument .27 .44 1122 NP arguments .23 .48 1093 NP arguments .29 .41 21
1 NP/1 PP argument .28 .48 58
Table 13: Probability of subject and NP complement postposing in fourHomeric data sets
Language Change: lecture 4 24
Tuesday, July 8, 14
III. The evolution of VX and XV inOld French
Tuesday, July 8, 14
• In tensed sentences only non-finite VPs are considered to avoid interference from Vfin-to-C and Vfin-to-T movement.
• Sentences in which a target clausal complement or adjunct moves further left than T are excluded since their “underlying” position relative to the non-finite verb is not recoverable.
• Sentences in which a target complement or adjunct is a clitic or empty category are excluded since their position is fixed or undeterminable.
Preliminaries, I
Tuesday, July 8, 14
Preliminaries, II
Old French texts contain occasional examples of OVT(ense) word order, superficially similar to the word order in German subordinate clauses:
(1) ...weil Maria das Brot gegessen hat because Maria the bread eaten has
(2) Quant l' ostes ce escouté eut when the army that heard had (1190-BORON-R,38.579)
Tuesday, July 8, 14
Preliminaries, III
However, in Old French, OVT(ense) word order is always due to leftward movement of a VP to an A-bar position. This conclusion is supported by two facts:
•Pre-Tense VPs always occur to left of any clitics associated with the finite verb.
•Pre-Tense VPs may be either OV or VO, superficially violating in the Final-Over-Final Constraint.
Tuesday, July 8, 14
(1) des que vos tant dit m' an avez since that you so-much told me of-it have (1170-YVAIN-R,151.5230)
(2) Seignur servir bien deit l'um tel lord serve well owes one such (1120-BRENDAN-R,55.666)
(3) Ainz que trovét nule rien ait. before that found any thing has (1120-BRENDAN-R,70.1085)
Tuesday, July 8, 14
(1) (Subject)> Finite V > Object
Nostres Sires savoit tout bien 'Our Lord knew everything well.' (1190-BORON-R,9.110)
si avroiz molt grant aventage 'So you would have a very great advantage.' (1170-YVAIN-R,41.1361)
Excluded finite clause types 1
Tuesday, July 8, 14
(2) (Subject)> Object > Finite V
Uns viels prestre la porte garde 'An old priest was guarding the door' (116X-MARIE-DE-FRANCE-R,16.276)
et vos enor et joie rande 'and God give you honor and joy' (1170-YVAIN-R,162.5687)
Excluded finite clause types 2
Tuesday, July 8, 14
(1) Object>(Subject)>Finite V> Nonfinite V
Sa grant honur a grant dol ad turnede (f. sg.) 'He has turned his great honor to great sorrow.' (10XX-ALEXIS-V,29.282)
(2) (Subject)>Object>Finite V>Nonfinite V
Li amiralz .X.escheles ad justedes (f. pl.) 'The admiral arranged ten batallions.' (1100-ROLAND-V,234.3228)
Excluded finite clause types 3
Tuesday, July 8, 14
(1) Rollant ad mis l’ olifan a sa buche 'Roland raised the ivory horn to his mouth.' (1100-ROLAND-V,133.1772)
(2) Li reis Marsilie out sun cunseill finet 'King Marsilla had adjourned his council.' (1100-ROLAND-V,5.53)
VO & OV word order: avoir+participle
Tuesday, July 8, 14
(1) Je veul avoir mon loier 'I want to have my pay.' (127X-CASSIDORUS-P,164.1546)
(2) Kar ne poeit le jur choisir 'For he cannot choose the day.' (116X-MARIE-DE-FRANCE-R,111.2262)
VO & OV word order: modal+infinitive
Tuesday, July 8, 14
(1) é pursievre David cessad 'and he ceased to pursue David' (1150-QUATRELIVRE-P,47.1793)
(2) Le abét e tuz baiser enprent 'He begins to kiss the abbot and everyone.' (1120-BRENDAN-R,47.464)
VO & OV word order: other nonfinite clauses
Tuesday, July 8, 14
OV decline with multi-word quantified and non-quantified objects
Tuesday, July 8, 14
OV decline by multi-word object length, quantified and non-quantified combined
EOF
Tuesday, July 8, 14
OV decline with one word quantified and non-quantified objects
EOF
Tuesday, July 8, 14
Decline of OV word order by clause type
EOF
Tuesday, July 8, 14
(1) Li emperere ad prise sa herberge (f. sg.) 'The emperor has taken his lodging.' (1100-ROLAND-V,182.2486)
(2) Vos li avez tuz ses castels toluz (m. pl.) 'You have taken all his castles from him.' (1100-ROLAND-V,16.208)
avoir+participle with object agreement
Tuesday, July 8, 14
(1) Li nostre deu i unt fait felonie (m. sg. - f. sg.) 'Our gods have committed a felony there.' (1100-ROLAND-V,188.2595)
(2) ... avoit tous les autres vaincu (m. sg. - m. pl.) '... (the emperor) had vanquished all the others.' (137X-PRISE-R,.639)
avoir+participle with object disagreement
Tuesday, July 8, 14
(1) Forment l’(= la prison) ont fermee et serree (f. sg.) 'They closed and sealed it securely.' (1190-BORON-R,25.398)
(2) tuz les i ad perdut (m. pl. - m. sg.) 'He lost them all there.' (1100-ROLAND-V,152.2053)
avoir+participle with agreeing and disagreeing clitic objects
Tuesday, July 8, 14
Object-participle agreement by DP type
EOF
Tuesday, July 8, 14
(1) s' avoient les espees traites desos les capes (f. pl.) 'So they had their swords drawn under their capes.' (122X-AUCASSIN-14-P,.21)
(2) N' ot drap vestu fors la chemise. 'He had no clothes on except for his shirt.' (116X-MARIE-DE-FRANCE-R,99.2003)
avoir + participial small clause
Tuesday, July 8, 14
(1) out vestue sa brunie (f. sg.) 'He had put on his body armor.' (1100-ROLAND-V,29.364)
(2) s' avoient bien lïez de cordes les piez (m. pl. ) 'So they had their feet tightly bound with cords.' (1170-YVAIN-R,125.4314)
Possible cases of avoir + participial small clause with postposed DP
Tuesday, July 8, 14
Decline of OV word order in clauses with avoir+participle
EOF
Tuesday, July 8, 14
Decline of OV word order: agree-marked versus neutralized avoir+participle
EOF
Tuesday, July 8, 14
Decline of OV word order by clause type: all avoir+participle versus others
EOF
Tuesday, July 8, 14
(1) En celui temps Bruthus avoit congneue charnelment Ynogen sa femme (f. sg.) 'During this time Brutus had had intercourse with his wife Ynogen.' (133X-PERCEFOREST-P,87.443)
(2) il n' avoit pas mis encores son consel ensamble 'He had not yet assembled his council.' (1373-FROISSART-P,402.8059)
Cases of raising of a nonfinite verb across an adverb
Tuesday, July 8, 14
Frequency of the raising of nonfinite verbs over adverbs by date
EOF
Tuesday, July 8, 14
Decline of OV word order by clause type: all avoir+participle versus others
EOF
Tuesday, July 8, 14
V > DO DO > V Row Totals Estimated DO scrambling
V > IO 40 2 42 .048
IO > V 5 5 10
Column Totals 45 7 52
Estimated IO scrambling .11
VO and OV word order in sentences with full DP direct and indirect objects in Early Old French.
Expected DO,IO > V order = .11*.048*52 = .28
Observed DO,IO > V order = 5 Chi-square =81.14
V > DO DO > V Row Totals Estimated DO scrambling
V > IO 41 3 44 .068
IO > V 6 1 7
Column Totals 47 4 51
Estimated IO scrambling .13
VO and OV word order in sentences with full DP direct and indirect objects in Later Old French.
1
V > DO DO > V Row Totals Estimated DO scrambling
V > IO 40 2 42 .048
IO > V 5 5 10
Column Totals 45 7 52
Estimated IO scrambling .11
VO and OV word order in sentences with full DP direct and indirect objects in Early Old French.
Expected DO,IO > V order = .11*.048*52 = .28
Observed DO,IO > V order = 5 Chi-square =81.14
V > DO DO > V Row Totals Estimated DO scrambling
V > IO 41 3 44 .068
IO > V 6 1 7
Column Totals 47 4 51
Estimated IO scrambling .13
VO and OV word order in sentences with full DP direct and indirect objects in Later Old French.
1
VX ︎ ↔︎ XV word orders in clauses with full DP direct and indirect objects in Early Old French
Tuesday, July 8, 14
V > DO DO > V Row totals Estimated DO scrambling
V > PP 237 54 291 .19
PP > V 24 38 62
Column totals 261 92 353
Estimated PP scrambling .092
VO and OV word order in sentences with full DP direct object and a clausal PP in Early Old French.
Expected DO,PP > V order = .092*.19*353 = 6.0
Observed DO,PP > V order = 38 Chi-square = 169.8
V > DO DO > V Row totals Estimated DO scrambling
V > PP 200 58 258 .22
PP > V 3 3 6
Column totals 203 61 264
Estimated PP scrambling .015
VO and OV word order in sentences with full DP direct object and a clausal PP in Later Old French.
Expected DO,PP > V order = .015*.22*264 = .88
Observed DO,PP > V order = 3 Chi-square = 5.21
2
V > DO DO > V Row totals Estimated DO scrambling
V > PP 237 54 291 .19
PP > V 24 38 62
Column totals 261 92 353
Estimated PP scrambling .092
VO and OV word order in sentences with full DP direct object and a clausal PP in Early Old French.
Expected DO,PP > V order = .092*.19*353 = 6.0
Observed DO,PP > V order = 38 Chi-square = 169.8
V > DO DO > V Row totals Estimated DO scrambling
V > PP 200 58 258 .22
PP > V 3 3 6
Column totals 203 61 264
Estimated PP scrambling .015
VO and OV word order in sentences with full DP direct object and a clausal PP in Later Old French.
Expected DO,PP > V order = .015*.22*264 = .88
Observed DO,PP > V order = 3 Chi-square = 5.21
2
VX ︎ ↔︎ XV word orders in clauses with a full DP direct object and a clausal level PP in Early Old French
Tuesday, July 8, 14
V > DO DO > V Row totals DO scrambling rate
V > IO 40 2 42 .048
IO > V 5 5 10
Column totals 45 7 52
IO scrambling rate .11
VO and OV word order in sentences with full DP direct and indirect objects in Early Old French.
Expected DO,IO > V order = .11*.048*52 = .28
Observed DO,IO > V order = 5 Chi-square = 81.14
V > DO DO > V Row totals Estimated DO scrambling
V > IO 41 3 44 .068
IO > V 6 1 7
Column totals 47 4 51
Estimated IO scrambling .13
VO and OV word order in sentences with full DP direct and indirect objects in Later Old French.
Expected DO,IO > V order = .13*.068*51 = .44
Observed DO,IO > V order = 1 Chi-square = 0.6966
1
V > DO DO > V Row totals DO scrambling rate
V > IO 40 2 42 .048
IO > V 5 5 10
Column totals 45 7 52
IO scrambling rate .11
VO and OV word order in sentences with full DP direct and indirect objects in Early Old French.
Expected DO,IO > V order = .11*.048*52 = .28
Observed DO,IO > V order = 5 Chi-square = 81.14
V > DO DO > V Row totals Estimated DO scrambling
V > IO 41 3 44 .068
IO > V 6 1 7
Column totals 47 4 51
Estimated IO scrambling .13
VO and OV word order in sentences with full DP direct and indirect objects in Later Old French.
Expected DO,IO > V order = .13*.068*51 = .44
Observed DO,IO > V order = 1 Chi-square = 0.6966
1
VX ︎ ↔︎ XV word orders in clauses with full DP direct and indirect objects in later Old French
Tuesday, July 8, 14
V > DO DO > V Row totals Estimated DO scrambling
V > PP 237 54 291 .19
PP > V 24 38 62
Column totals 261 92 353
Estimated PP scrambling .092
VO and OV word order in sentences with full DP direct object and a clausal PP in Early Old French.
Expected DO,PP > V order = .092*.19*353 = 6.0
Observed DO,PP > V order = 38 Chi-square = 169.8
V > DO DO > V Row totals Estimated DO scrambling
V > PP 200 58 258 .22
PP > V 3 3 6
Column totals 203 61 264
Estimated PP scrambling .015
VO and OV word order in sentences with full DP direct object and a clausal PP in Later Old French.
Expected DO,PP > V order = .015*.22*264 = .88
Observed DO,PP > V order = 3 Chi-square = 5.21
2
V > DO DO > V Row totals Estimated DO scrambling
V > PP 237 54 291 .19
PP > V 24 38 62
Column totals 261 92 353
Estimated PP scrambling .092
VO and OV word order in sentences with full DP direct object and a clausal PP in Early Old French.
Expected DO,PP > V order = .092*.19*353 = 6.0
Observed DO,PP > V order = 38 Chi-square = 169.8
V > DO DO > V Row totals Estimated DO scrambling
V > PP 200 58 258 .22
PP > V 3 3 6
Column totals 203 61 264
Estimated PP scrambling .015
VO and OV word order in sentences with full DP direct object and a clausal PP in Later Old French.
Expected DO,PP > V order = .015*.22*264 = .88
Observed DO,PP > V order = 3 Chi-square = 5.21
2
VX ︎ ↔︎ XV word orders in clauses with a full DP direct object and a clausal level PP in later Old French
Tuesday, July 8, 14
Finis
Tuesday, July 8, 14